Futuresony commited on
Commit
b78b527
·
verified ·
1 Parent(s): e2fbab6

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +8 -13
app.py CHANGED
@@ -3,29 +3,24 @@ from huggingface_hub import InferenceClient
3
 
4
  client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf")
5
 
6
- def format_clean_prompt(user_input):
7
- """Formats input without extra labels"""
8
- return f"{user_input}\n"
9
-
10
- def respond(message, history, system_message, max_tokens, temperature, top_p):
11
- formatted_prompt = format_clean_prompt(message)
12
 
13
  response = client.text_generation(
14
  formatted_prompt,
15
- max_new_tokens=max_tokens,
16
- temperature=temperature,
17
- top_p=top_p,
18
  )
19
 
20
- # ✅ Clean output by removing extra newlines
21
- yield response.strip()
22
 
23
  demo = gr.ChatInterface(
24
  respond,
25
  additional_inputs=[
26
- gr.Slider(minimum=1, maximum=250, value=128, step=1, label="Max new tokens"),
27
  gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
28
- gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p (nucleus sampling)"),
29
  ],
30
  )
31
 
 
3
 
4
  client = InferenceClient("Futuresony/future_ai_12_10_2024.gguf")
5
 
6
+ def respond(message, history, max_tokens, temperature, top_p):
7
+ formatted_prompt = message.strip() # Ensure clean input
 
 
 
 
8
 
9
  response = client.text_generation(
10
  formatted_prompt,
11
+ max_new_tokens=int(max_tokens), # ✅ Ensure integer
12
+ temperature=float(temperature),
13
+ top_p=float(top_p),
14
  )
15
 
16
+ yield response.strip() # ✅ Clean output
 
17
 
18
  demo = gr.ChatInterface(
19
  respond,
20
  additional_inputs=[
21
+ gr.Slider(minimum=1, maximum=250, value=128, step=1, label="Max new tokens"), # ✅ Ensure integer input
22
  gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
23
+ gr.Slider(minimum=0.1, maximum=1.0, value=0.95, step=0.05, label="Top-p"),
24
  ],
25
  )
26